You will not just manage a feature; you will own the fusion of these technologies. From Generative Video models that render lifelike avatars, to Speech AI engines (ASR/TTS) that handle nuanced voice interaction, to the LLM & RAG layers that provide the brain and context-you will lead the strategy that turns these individual components into a unified, low-latency Face-to-Face conversational engine for the enterprise.
The day-to-day
Strategic Collaboration: Partner closely with the Head of AI and Head of Product to align the generative media roadmap with technical constraints.
Orchestrate the Multi-Model Stack: Define the product vision for a complex architecture-integrating Latent Diffusion for visuals, Transformer-based Speech AI, and LLMs-into a seamless real-time loop optimized for minimal Time-to-First-Token (TTFT).
Advance Video Generation Tech: Spearhead the deployment of diffusion-based video synthesis and neural rendering.
Establish Benchmarks: Set standards for temporal consistency, high-fidelity lip-sync, and dynamic avatar behaviors that meet enterprise bitrate standards.
Optimize the Brain (LLM & RAG): Lead the strategy for our cognitive layer, overseeing Vector Database selection and embedding models for RAG pipelines.
Ensure Accuracy: Ground AI responses in enterprise data, strictly minimizing hallucination rates and optimizing context window usage.
5+ years of experience in product management or a related technical field.
MSc or PhD in a Machine Learning field (or equivalent advanced technical degree).
A visionary capable of managing a holistic multi-model ecosystem where Video Generation, Speech AI, and LLM/RAG architectures intersect.
Ability to balance SOTA research with inference COGS and GPU availability to drive scalable business growth.
Deep understanding of the fusion between Generative Video models, Speech AI engines, and LLM layers.
